Presented herein is the first morphological analysis of turtle relationships to examine the monophyly of many turtle groups by using only single species as terminals and by integrating a large number of primitive fossil taxa. The data matrix consists of 136 osteological parsimony informative characters with 169 derived character states for 45 fossil and 22 living species of the clade Testudinata. The results corroborate the monophyly of a large number of previously hypothesized clades, but refute the accepted hypothesis regarding the basal split of living turtles. In particular, the primitive turtles Proterochersis robusta, Kayentachelys aprix, Mongolochelys efremovi, Meiolania platyceps, and Kallokibotion bajazidi are removed from their current position as crown turtles and placed along the phylogenetic stem of this clade. The age of the turtle crown is thereby adjusted from the Late Triassic to the Late Jurassic, which is relevant to testing molecular clock hypotheses. This revised topology has important implications for the evolution of several character complexes, because it implies that the common ancestor of all living turtles must have had a partially braced brain case and a primitive trochlear mechanism. Other noteworthy conclusions include the tentative exclusion of protostegids from Chelonioidea, the placement of Platysternon megacephalum outside of Chelydridae, and the tentative interpretation of Sandownia harrisi as a basal eucryptodire.

No abstract

No abstract

Competing hypotheses of early turtle evolution contrast sharply in implying very different ecological settings-aquatic versus terrestrial-for the origin of turtles. We investigate the palaeoecology of extinct turtles by first demonstrating that the forelimbs of extant turtles faithfully reflect habitat preferences, with short-handed turtles being terrestrial and long-handed turtles being aquatic. We apply this metric to the two successive outgroups to all living turtles with forelimbs preserved, Proganochelys quenstedti and Palaeochersis talampayensis, to discover that these earliest turtle outgroups were decidedly terrestrial. We then plot the observed distribution of aquatic versus terrestrial habits among living turtles onto their hypothesized phylogenies. Both lines of evidence indicate that although the common ancestor of all living turtles was aquatic, the earliest turtles clearly lived in a terrestrial environment. Additional anatomical and sedimentological evidence favours these conclusions. The freshwater aquatic habitat preference so characteristic of living turtles cannot, consequently, be taken as positive evidence for an aquatic origin of turtles, but must rather be considered a convergence relative to other aquatic amniotes, including the marine sauropterygians to which turtles have sometimes been allied.